CN112714037A - Method, device and equipment for evaluating guarantee performance of online service quality - Google Patents

Abstract

The invention provides a method, a device and equipment for evaluating guarantee performance of on-line service quality, wherein the method comprises the following steps: analyzing the network performance to obtain the service capability and the load of the system; acquiring the time delay upper bound of the system according to the service capacity and the load; and evaluating the quality of the online service according to the upper time delay bound, wherein the invention aims to provide an evaluation result for the online service with burst traffic so as to optimize the service capability of the system.

Description

An online service quality assurance performance evaluation method, device and equipment

technical field

The invention relates to the field of network performance evaluation, in particular to a method, device and equipment for ensuring performance evaluation of online service quality.

Background technique

In the research based on online network traffic performance analysis, some scholars have carried out research on network performance analysis from the research results of Web service quality analysis. Aiming at the problem that the unstable network environment leads to the existence of noisy data in the Web service quality (QoS) data, which reduces the prediction accuracy of Web service quality, a hybrid collaborative filtering Web service quality value prediction method based on Bayesian classification is proposed. This method uses the Bayesian algorithm to classify the Web service quality data and obtains the probability of each classification, uses the classification results to determine the possible range of missing values, and filters the similar neighbors of users and services. By introducing the classification probability, The traditional collaborative filtering method is improved to obtain the final missing value prediction result, which eliminates the influence of noise data on the prediction of Web service quality to a certain extent. According to the characteristics of cloud services, the service-oriented architecture and relative weights of evaluation indicators such as service performance, reliability, throughput, and utilization of network services are used to analyze the network service quality under cloud services. Aiming at the traffic scenarios of large-scale distributed e-commerce clusters and the needs of dynamic capacity planning, Shi Jingwen et al. proposed a real-time traffic forecasting framework including uncertainty estimation. The framework is based on multivariate long-short-term memory network autoencoders and Bayesian theory, which can give accurate uncertainty interval estimates while making deterministic predictions of traffic. Based on Amazon's EC2, Iosup and others designed a cloud service performance evaluation model using scientific computing methods. The model mainly analyzed the computing performance, load and energy consumption of the cloud computing platform;

However, in the prior art, and in order to analyze the service quality of the online teaching network under emergencies, and to provide a relatively reliable evaluation result, the present application is made in view of this.

SUMMARY OF THE INVENTION

The invention discloses an online service quality assurance performance evaluation method, device and equipment, which aim to provide an evaluation result for online services with burst traffic, so as to optimize the service capability of the system.

The first embodiment of the present invention provides an online service quality assurance performance evaluation method, including:

Analyze network performance to obtain system service capability and system load;

Obtain the upper bound of the system delay according to the service capability and load;

According to the upper bound of the delay, the online service quality is evaluated.

Preferably, the analyzing network performance includes: analyzing blocking probability, analyzing immediate service probability, analyzing delay time, analyzing service concurrency capability, and analyzing load.

Preferably, according to the service capability and load, the acquisition of the upper bound of the delay of the system is specifically:

The service capability is β _e2e , the load is a(t), and the upper bound of the delay D _max The upper bound of the delay of the system is obtained by the following formula:

Formula 1:

Formula 2: D _max =sup{inf{τ≥0:α(s)≤β _e2e (s+τ)}};

Formula 3:

Among them, R _n1 is the service rate provided by the forward network service, R _n2 is the service rate provided by the backward network service, R _c is the service rate provided by the cloud service, and f is the impact of the cloud service on data transmission when processing the computing process factor.

Preferably, the online service quality is evaluated, specifically:

Calculating the formulas 1, 2, and 3 to obtain:

Wherein, R _e2e =min{R _n1 ,R _c ,R _n2 /f}, T _e2e =T _n1 +T _c +T _n2 ;

When the load is α(t)=M+pt, the upper bound of the system delay is D _max =T _e2e +M/R _e2e =T _c +T _e +M/R _e2e ;

Wherein, T _n =T _n1 +T _n2 ;

Preferably, it also includes: in the case of ignoring the signal processing delay, the service delay parameter T is the sum of the link transmission delay and the packet processing delay, namely T=L/R+L/C, where L represents the maximum packet long, C represents the minimum link transmission rate, and R represents the number of user requests.

Wherein, T _n =L(1/R _n1 +1/R _n2 +2/C);

The upper bound of the delay of the system is D _max =T _c +L(1/R _n1 +1/R _n2 +2/C)+M/R _e2e ; where, R _e2e =min{R _n1 ,R _c ,R _n2 /f}.

The second embodiment of the present invention provides an online service quality assurance performance evaluation device, including:

The network performance analysis module is used to analyze the network performance and obtain the service capability of the system and the load of the system;

a delay upper bound obtaining module, used for obtaining the system delay upper bound according to the service capability and load;

The online service quality evaluation module is used to evaluate the online service quality according to the upper bound of the delay.

Preferably, the delay upper bound obtaining module is specifically used for:

The service capability is β _e2e , the load is a(t), and the upper bound of the delay D _max The upper bound of the delay of the system is obtained by the following formula:

Formula 1:

Formula 2: D _max =sup{inf{τ≥0:α(s)≤β _e2e (s+τ)}};

Formula 3:

Among them, R _n1 is the service rate provided by the forward network service, R _n2 is the service rate provided by the backward network service, R _c is the service rate provided by the cloud service, and f is the impact of the cloud service on data transmission when processing the computing process factor.

Preferably, the online service quality assessment module is specifically used for:

Calculating the formulas 1, 2, and 3 to obtain:

Wherein, R _e2e =min{R _n1 ,R _c ,R _n2 /f}, T _e2e =T _n1 +T _c +T _n2 ;

When the load is α(t)=M+pt, the upper bound of the system delay is D _max =T _e2e +M/R _e2e =T _c +T _e +M/R _e2e ;

Wherein, T _n =T _n1 +T _n2 ;

Preferably, the online service quality evaluation module is further configured to: in the case of ignoring the signal processing delay, the service delay parameter T is the sum of the link transmission delay and the packet processing delay, that is, T=L/R+ L/C, where L represents the maximum packet length, C represents the minimum link transmission rate, and R represents the number of user requests;

Wherein, T _n =L(1/R _n1 +1/R _n2 +2/C);

The upper bound of the delay of the system is D _max =T _c +L(1/R _n1 +1/R _n2 +2/C)+M/R _e2e ; where, R _e2e =min{R _n1 ,R _c ,R _n2 /f}.

A third embodiment of the present invention provides an online service quality assurance performance evaluation device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, the processor executing The computer program implements an online service quality assurance performance evaluation method described in any one of the above.

Based on an online service quality assurance performance evaluation method, device and equipment provided by the present invention, by analyzing the performance of the network, the service capability of the system and the load of the system are obtained, and the delay is proposed for the service capability and load obtained by the analysis. The upper bound is obtained by analyzing different factors of the upper bound of delay to obtain an online service quality evaluation report with high reliability.

Description of drawings

FIG. 1 is a schematic flowchart of a method for evaluating the performance of online service quality assurance provided by the first embodiment of the present invention;

2 is a schematic diagram of a multi-service station hybrid queuing model provided by an embodiment of the present invention;

3 to 10 are simulation experiment diagrams provided by an embodiment of the present invention;

FIG. 11 is a schematic structural diagram of an online service quality assurance performance evaluation module provided by the second embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

In order to better understand the technical solutions of the present invention, the embodiments of the present invention are described in detail below with reference to the accompanying drawings.

It should be understood that the described embodiments are only some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. As used in the embodiments of the present invention and the appended claims, the singular forms "a," "the," and "the" are intended to include the plural forms as well, unless the context clearly dictates otherwise.

It should be understood that the term "and/or" used in this document is only an association relationship to describe the associated objects, indicating that there may be three kinds of relationships, for example, A and/or B, which may indicate that A exists alone, and A and B exist at the same time. B, there are three cases of B alone. In addition, the character "/" in this document generally indicates that the related objects are an "or" relationship.

Depending on the context, the word "if" as used herein can be interpreted as "at" or "when" or "in response to determining" or "in response to detecting." Similarly, the phrases "if determined" or "if detected (the stated condition or event)" can be interpreted as "when determined" or "in response to determining" or "when detected (the stated condition or event)," depending on the context )" or "in response to detection (a stated condition or event)".

The "first\second" mentioned in the embodiment is only to distinguish similar objects, and does not represent a specific order for the objects. It is understood that "first\second" can be interchanged with specific order or sequence. It should be understood that the "first\second" distinctions may be interchanged under appropriate circumstances to enable the embodiments described herein to be practiced in sequences other than those illustrated or described herein.

The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

The invention discloses an online service quality assurance performance evaluation method, device and equipment, which aim to provide an evaluation result for online services with burst traffic, so as to optimize the service capability of the system.

Referring to FIG. 1, the first embodiment of the present invention provides an online service quality assurance performance evaluation method, which can be performed by an online service quality assurance performance evaluation device (hereinafter referred to as evaluation device), in particular, Executed by one or more processors within the evaluation device to implement the following steps:

S101, analyze the network performance to obtain the service capability of the system and the load of the system;

In this embodiment, the burst traffic may be analyzed first, in particular, the burst online teaching traffic may be analyzed;

With a directed connected graph N=<V,E>, denoted n=|V|, m=|E|, each edge <i, j> has a non-negative number C(i, j), called edge <i , j> the statistical flow. N has two special vertices s and t, s is called the sending point, t is called the receiving point, and the remaining vertices are called the intermediate points [11]. Call N the aggregated traffic at the network egress, and denote it as N=<V,E,c,s,t>.

Let f:E→R*, where R* is the set of non-negative numbers that satisfy the following conditions:

(1) Flow limit

(2) Load balancing

Let f be a feasible flow on the aggregated flow N, and call the net flow of the delivery point s the flow of f, denoted as v(f), that is

A feasible flow with the largest change in flow value is called a burst flow.

To analyze the aggregated stream:

且

则Suppose N=<V,E,c,s,t>,

and

but

From the above one can get:

是任一割集，则

If f is any aggregated flow on N,

is any cut set, then

是任一割集，且

则是限制最大流，

是最小割集。If f is any aggregated flow on network N,

is any cut set, and

is to limit the maximum flow,

is the minimum cut set.

Assuming that there is at most one edge between each vertex in N, if there are two edges <i,j> and <j,i> between i and j, a vertex k can be inserted on <j,i>, and < j,i> is divided into two sides <j,k> and <k,i>, and the capacity is equal to c(j,i). So:

maxv(f)=s.t.f(i,j)≤c(i,j),<i,j>∈E (9)

f(i,j)≥0,<i,j>∈E (12)

v(f)≥0 (13)

Analyze the blocking probability:

When a large number of users cannot enter the queue buffer due to burst traffic, network congestion will occur. The blocking probability [12] is an important indicator to measure the network QOS. For waiting users, the system will adopt a first-come, first-served policy. After the user enters the buffer, there may be two different results: queuing or being rejected. Assuming that the position of a single user R in the queue conforms to a normal distribution, the probability of R in batch j is:

Let the queue length of R be and the number of requests be, then, the subsequent users cannot enter the waiting area, that is, there will be a block of requests in the batch. Its blocking probability is

Analysis of immediate service probability:

When the server requested by R is idle, the service is immediately available. The queue length of the current system is i (i<m), and the number of batch arrivals is j. When j≤m-i, all requests are immediately served; when j>m-i, the system adopts partial acceptance strategy and first-come-first-served order, so a request at position [1,j-m+i] can be serviced immediately. Therefore, the immediate service probability [13] is

Analyze the delay time:

User delay time is the sum of system service and waiting time. According to Equation 14, W _q (t) is the probability distribution function of the waiting time t. When the waiting time is 0, its probability is:

If the position of R is n,n∈[1,j], because the service cannot be obtained immediately, according to the principle of first-come, first-come, it is necessary to wait for the completion of the previous service request before obtaining the service. Assuming that the number of requests in R is l, m is the number of service desks, and the outgoing flow of l requests conforms to the l-order Erlang distribution of mμ. There are two cases of L:

为When i < m, the R request arrives. When j ∈ (mi, Ni], when n ∈ (mj, j], R enters the queue to wait. When j ∈ (Ni, ∞], and n ∈ (mj, Nj], it also waits. There is l =n-m+i, the distribution function of the waiting time

for

为When i ≥ m, there are no idle service desks when the request arrives. When j∈[1,Ni], n∈[1,j], the request is queued. When j∈(Ni,∞], and n∈[1,Ni], it also waits. There is l=n-m+i, the distribution function of waiting time

for

Therefore, the waiting time W _q (t) is

和

可分别由式(18)和式(19)代入。延时时间W＝W_q+W_s， W_q和W_s互相独立。根据分布函数的卷积公式，可以得到延时时间为：in,

and

can be substituted by formula (18) and formula (19), respectively. Delay time W=W _q +W _s , W _q and W _s are independent of each other. According to the convolution formula of the distribution function, the delay time can be obtained as:

Analyze the service concurrency capability:

In cloud service systems, user requests often arrive in batches (eg, users need to deploy multiple VMs). In this paper, according to the literature [14-15], starting from the characteristics of user batch requests and traffic characteristics, the M ^x /M/m/m+r queuing model will be modeled and the service performance of cloud computing will be analyzed. According to the common performance indicators in the queuing system, the expression of the corresponding indicators in the batch queuing is given, and the service quality of the cloud service system is analyzed to ensure QoS and avoid violating the service level agreement (SLA).

x＝1,2,…,k(k 为正整数)。When the business system is in normal service, there are usually a large number of users requesting to enter the service desk. When a user sends a request, the service desk will adaptively provide different types of services according to the user's needs, and the user will also queue up into the buffer. The queuing model M ^x /M/m/m+r is a multi-server hybrid queuing model, which belongs to a finite state Markov process, as shown in Figure 2. The working characteristics of the model are: data streams arrive at the service desk in batches, λ is the arrival time interval of the data stream obeying Poisson flow; the number of requests in each batch is a random variable x, and its probability distribution is P(X=x)=c _i ,

x=1,2,...,k (k is a positive integer).

已知当ρ<1时，系统存在平稳状态，状态转移满足如下规则。Set the queue length i of the number of user requests in the system as the state variable of the system, its probability is π _i , the system capacity N=m+r, the traffic intensity

It is known that when ρ<1, the system has a stationary state, and the state transition satisfies the following rules.

①For the transition out of state i, starting from any state i, it can reach any subsequent state i+1, i+2,..., or i+k in one step to the right, that is, in the i state there are Requests with batch x arrive at the system. And the left can only reach its adjacent state i-1, that is, at the same time, only one request is completed and left the system.

②When i is transferred in, the left state of state i at this time is i-1, i-2, ..., i-k arrives directly, that is, after the request with batch x reaches the system, its queue length is i. The state on the right is i+1, i+2,..., i+k, when the adjacent state i+1 can reach the i state.

③ After the service ends, the transition process from state i to i-1 state, the transition probability has different changing laws according to the value of parameter i. Taking the state m as the demarcation point, for the state i (0<i<m) on the left side of m, the transition probability is iμ, and the transition probability on the right side is mμ.

According to the Chapman-Kolmogorov-equation [], it can be obtained:

可获得所有队列长度的概率分布。From equation (21), π _i+1 can be recursively derived from π _i , and the relationship between π ₁ , π ₂ ,...,π _N and π ₀ can be obtained by

Probability distributions for all queue lengths are available.

Through the above analysis, we can conclude that increasing the number of service desks is conducive to improving service performance. In the face of a large burst of user requests, we need to increase the terminal server (virtual machine or host) in order to effectively ensure the user's QoS.

Analyze the load:

When the integrated service system provides the same service capability, the service performance obtained by the user will be different due to the difference in its load. Therefore, in order to more accurately analyze the service performance of the cloud service integrated service system, it is necessary to propose a The method is used to describe the load of the integrated service system, that is, the data flow of the end user accessing the integrated service system.

The data flow of the user's access to the service system is limited by the access capability of the forward network service system. The forward network service system needs to detect whether the data flow reaching the entrance meets the conditions. When the conditions are not met, the network needs to make corresponding adjustments. Therefore, the data flow of the end user accessing the forward network service system is the load of the integrated service system.

Since most network systems use network traffic control policies to constrain the arriving aggregated flow, and in order to express the characteristics of the flow, in practical applications, load balancing is to schedule the jobs submitted by users to different virtual machine resources, so that the Work load is shared among virtual machines or hosts in the system to complete job execution. Of course, the existing resource pool can have tens or hundreds of thousands of virtual machines and service nodes of the host. If the load is not balanced, it is very likely that all services will be interrupted. Therefore, this paper assumes that the load of the online teaching service system can be expressed in the form of a(t)=M+pt, where M represents the maximum burst volume, p represents the arrival rate, and p must satisfy p≤R, otherwise the delay limit will tend to infinity.

To analyze the limited burst model:

Through the analysis of network traffic indicators in online teaching, the burst limiting model first considers the setting that affects the burstiness of large traffic, which is called "burst limit". We use a non-negative number R to represent a service flow on a given communication link. For any two moments X and y, and y≥x, the integral of R over the interval [x, y] indicates that the service is charged in time The amount of data transmitted on the link within [x, y]. R(t)[0, 1] is expressed as the instantaneous flow of online services at time t. When R(t) is 0, it means that its link is idle; when R(t) is 1, it means that its link is fully loaded for transmission. Given σ≥0 and ρ≥0, R～(σ,ρ), and when y≥x is satisfied for all x and y, then:

In the formula, ρ is the average rate, σ is the burst limit, for a fixed value ρ, the larger the σ, the greater the burst capability.

When the traffic is defined as the upper bound burst traffic, there are:

In the formula, a is the function attenuation factor. When there is an upper bound on the total flow rate in any time period [x, y], the upper bound is a decreasing exponential function, ρ is the upper bound flow rate, Ae ^-aσ is the limit function. If traffic is limited by this exponential function, then network latency and queue lengths are distributed exponentially decaying.

For random restricted emergencies, when f(σ)∈F, given σ>0, 0≤x<y, then we have:

ρ is the upper bound flow rate, and f(σ) is the limiting function.

S102, according to the service capability and load, obtain the upper bound of the delay of the system; specifically:

Stable performance is a measure for users to choose online services. Latency is an important parameter in performance indicators. In practical application scenarios, service quality is directly affected by latency performance. For example, when students choose to conduct online cloud teaching, they hope that the selected cloud service can complete their request in the shortest possible time. The traditional time delay analysis methods can be divided into two categories: one is the statistical theoretical analysis method, and the other is the time series analysis method. However, both methods can only obtain the statistical value or approximate value of the average delay, and cannot obtain the upper bound of the delay that directly affects the quality of cloud services and user service selection. Therefore, this embodiment focuses on discussing the upper bound of the delay of the converged service system.

The service capability is β _e2e , the load is a(t), and the upper bound of the delay D _max The upper bound of the delay of the system is obtained by the following formula:

Formula 1:

Formula 2: D _max =sup{inf{τ≥0:α(s)≤β _e2e (s+τ)}};

The service capability curve of each service component in the cloud service system can be represented by the LR function. Therefore, this paper assumes that the service capability of each service component in the service system and the scaling curve of the scaling function S are respectively

Formula 3:

Among them, R _n1 is the service rate provided by the forward network service, R _n2 is the service rate provided by the backward network service, R _c is the service rate provided by the cloud service, and f is the impact of the cloud service on data transmission when processing the computing process factor.

S103: Evaluate the online service quality according to the upper bound of the delay. Specifically:

The end-to-end service capability of the integrated service system is obtained by calculating the formulas 1, 2, and 3 as follows:

Wherein, R _e2e =min{R _n1 ,R _c ,R _n2 /f}, T _e2e =T _n1 +T _c +T _n2 ;

It can be seen from the above formula that for the cloud service integrated service system, if each network service component is described as an LR function, and the cloud service is described as an LR function with a linear scaling curve, then the end-to-end service provided by the entire service system Capability can be described as an LR function constrained by a scaling curve. Its service delay parameter T is the sum of the delay parameters of each service component in the whole system. The service rate is the minimum value among the forward network service rate, the cloud service rate, and the transmission service rate of the backward network service.

When the load is α(t)=M+pt, the upper bound of the system delay is D _max =T _e2e +M/R _e2e =T _c +T _e +M/R _e2e ;

Wherein, T _n =T _n1 +T _n2 ;

In this embodiment, for network services, the delay parameter T in the LR function reflects the system properties of the network, and it can be regarded as a busy period of a network session, the first bit stream is transmitted in the worst case. required time. In the case of ignoring the signal processing delay, the service delay parameter T is the sum of the link transmission delay and the packet processing delay, that is, T=L/R+L/C, where L represents the maximum packet length and C represents the minimum chain is the transmission rate, and R represents the number of user requests.

Wherein, T _n =L(1/R _n1 +1/R _n2 +2/C);

The upper bound of the delay of the system is D _max =T _c +L(1/R _n1 +1/R _n2 +2/C)+M/R _e2e ; where, R _e2e =min{R _n1 ,R _c ,R _n2 /f}.

It can be seen from the above formula that when the network service rate is a constant value and the burst volume of the data stream is 0, the upper bound of the delay of the integrated service system is a constant value, independent of the arrival rate of the data stream. This is because this chapter assumes that the arrival rate is not greater than the system service rate, so that the data flow entering the system will be serviced immediately without causing latency in the system. At this time, the upper bound of the delay required in this paper is the processing time of the system.

Specifically, please refer to Figure 3 and Figure 4, the delays under different scaling factors and different IP numbers (that is, the number of users), it can be seen that when the network service rate is very small, the network delay will increase with the increase of the service rate. Increase. When f and K continue to increase, the delay will also increase with the service rate, and the upper bound of the delay will continue to decrease, and the traffic will increase the load of the backward transmission network after the traffic is processed by the server. Therefore, the network service rate and network delay will have an impact under different circumstances, and the degree of impact will be different.

We use different network node numbers to simulate the impact on network performance data. When f=0.5 and f=2, when the number of nodes is 50, 100, 500, and 1000, respectively, Figure 5 (f=0.5) and Figure 6 (f=2 ), the simulation results obtained for the service rate for different number of nodes are given. When the number of nodes in the network used is small, when the number of network nodes (n) is 50, the obtained network delay is relatively large, and the greater the service rate, the longer the delay increases; and when the number of network nodes (n) ) is 100 and 500, the delay shown in the figure is already much smaller than the delay of 50 network nodes. When n=1000, the delay is getting smaller and smaller, which means that the more nodes there are, the smaller the delay is.

We assume that the maximum burst size is 2000Mb, and the burst traffic of the forward load is consistent with the backward load. It can be seen from Figure 7 (f=0.5) and Figure 8 (f=2) that with the increase of the maximum burst size, the delay of the system also increases. This is because when the service rate is constant, the network delay of the system also increases. This is because when the service rate is constant, the system delay is only related to the maximum burst size, and the larger the maximum burst size, the longer the system processing time. In Fig. 8, when f=2, since the data passing through the server increases the load of the backward transmission network, and the system provides a certain service rate, the system delay will also increase.

Figure 9 (f=0.5) and Figure 10 (f=2) describe the performance comparison chart of the delay upper bound of the method in this paper and the prior art under different scaling factors and the same forward flow. It can be seen from the experiment that under the same service rate, when f=0.5, Rn=5000Mb/s and f=2, Rn=5000Mb/s, the evaluation method in this paper will obtain a smaller upper bound of delay. This is because the prior art cannot allocate traffic according to the service capability of the link, resulting in excessive traffic load on the data link, leading to network congestion, thereby affecting the upper bound of the delay. The main factor that determines the performance evaluation method for limiting the burst model affecting the upper bound of the delay is the burst traffic. Although the arrival rate is related, it has little effect on the delay.

It can be seen that the upper bound of the delay increases with the increase of the service rate. When the service rate is constant, burst traffic is the main factor affecting the upper bound of the delay. When the number of users continues to increase, the service capability provided by the backward network service has an important impact on the upper bound of the delay. When the number of nodes is more, the service capability is stronger and the network delay is smaller, which shows that the computing power of the node also plays a major role in the upper bound of the delay. Therefore, in the face of sudden traffic, in order to avoid network congestion, network operators should calculate the upper bound of the data backlog of service nodes according to their own service capabilities and user requests in order to ensure network service quality, so as to provide network service nodes. Provides a basis for the allocation of the buffer size. Secondly, the export bandwidth should be expanded as much as possible and the load balancing of multiple bandwidths should be implemented to ensure that the bandwidth will not be congested. Finally, network operators should use the network virtual batch technology to provide network resources in the form of services, and provide a variety of network services to meet the needs of business individuation and diversification.

Referring to FIG. 11 , a second embodiment of the present invention provides an online service quality assurance performance evaluation device, including:

The network performance analysis module 201 is used to analyze the network performance and obtain the service capability of the system and the load of the system;

a delay upper bound obtaining module 202, configured to obtain the system delay upper bound according to the service capability and load;

The online service quality evaluation module 203 is configured to evaluate the online service quality according to the upper bound of the delay.

Preferably, the delay upper bound obtaining module 202 is specifically used for:

The service capability is β _e2e , the load is a(t), and the upper bound of the delay D _max The upper bound of the delay of the system is obtained by the following formula:

Formula 1:

Formula 2: D _max =sup{inf{τ≥0:α(s)≤β _e2e (s+τ)}};

Formula 3:

Among them, R _n1 is the service rate provided by the forward network service, R _n2 is the service rate provided by the backward network service, R _c is the service rate provided by the cloud service, and f is the impact of the cloud service on data transmission when processing the computing process factor.

Preferably, the online service quality evaluation module 203 is specifically used for:

Calculating the formulas 1, 2, and 3 to obtain:

Wherein, R _e2e =min{R _n1 ,R _c ,R _n2 /f}, T _e2e =T _n1 +T _c +T _n2 ;

When the load is α(t)=M+pt, the upper bound of the system delay is D _max =T _e2e +M/R _e2e =T _c +T _e +M/R _e2e ;

Wherein, T _n =T _n1 +T _n2 ;

Preferably, the online service quality evaluation module 203 is further configured to: in the case of ignoring the signal processing delay, the service delay parameter T is the sum of the link transmission delay and the packet processing delay, that is, T=L/R +L/C, where L represents the maximum packet length, C represents the minimum link transmission rate, and R represents the number of user requests;

Wherein, T _n =L(1/R _n1 +1/R _n2 +2/C);

The upper bound of the delay of the system is D _max =T _c +L(1/R _n1 +1/R _n2 +2/C)+M/R _e2e ; where, R _e2e =min{R _n1 ,R _c ,R _n2 /f}.

A third embodiment of the present invention provides an online service quality assurance performance evaluation device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, the processor executing The computer program implements an online service quality assurance performance evaluation method described in any one of the above.

Based on an online service quality assurance performance evaluation method, device and equipment provided by the present invention, by analyzing the performance of the network, the service capability of the system and the load of the system are obtained, and the delay is proposed for the service capability and load obtained by the analysis. The upper bound is obtained by analyzing different factors of the upper bound of delay to obtain an online service quality evaluation report with high reliability.

A fourth embodiment of the present invention provides a readable storage medium, which is characterized in that a computer program is stored, and the computer program can be executed by a processor of a device where the storage medium is located, so as to implement any one of the above-mentioned methods. An online service quality assurance performance evaluation method.

Exemplarily, the computer programs described in the third and fourth embodiments of the present invention may be divided into one or more modules, the one or more modules are stored in the memory, and the The processor executes to accomplish the present invention. The one or more modules may be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program in the guarantee performance evaluation device for realizing an online quality of service . For example, the apparatus described in the second embodiment of the present invention.

The processor may be a central processing unit (Central Processing Unit, CPU), or other general-purpose processors, digital signal processors (Digital Signal Processors, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf processors Programmable Gate Array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc., the processor is the control center of the online service quality assurance performance evaluation method, and uses various interfaces and lines to connect the entire system. The implementation of each part of the guarantee performance evaluation method for online service quality.

The memory can be used to store the computer programs and/or modules, and the processor implements online services by running or executing the computer programs and/or modules stored in the memory and calling the data stored in the memory Various functions of quality assurance performance evaluation methods. The memory may mainly include a stored program area and a stored data area, wherein the stored program area can store an operating system, an application program required for at least one function (such as a sound playback function, a text conversion function, etc.), etc.; the stored data area can store Data (such as audio data, text message data, etc.) created according to the use of the mobile phone, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory such as hard disk, internal memory, plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, at least one magnetic disk storage device, flash memory device, or other volatile solid state storage device.

Wherein, if the implemented modules are implemented in the form of software functional units and sold or used as independent products, they may be stored in a computer-readable storage medium. Based on this understanding, the present invention can implement all or part of the processes in the methods of the above embodiments, and can also be completed by instructing relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, the computer When the program is executed by the processor, the steps of the foregoing method embodiments can be implemented. Wherein, the computer program includes computer program code, and the computer program code may be in the form of source code, object code, executable file or some intermediate form, and the like. The computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a U disk, a removable hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM, Read-Only Memory) , Random Access Memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable media may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer-readable media Electric carrier signals and telecommunication signals are not included.

It should be noted that the device embodiments described above are only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical unit, that is, it can be located in one place, or it can be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment. In addition, in the drawings of the apparatus embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art can understand and implement it without creative effort.

The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited to this. Substitutions should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims (10)

1. An online service quality assurance performance evaluation method is characterized in that, comprising: Analyze network performance to obtain system service capability and system load; Obtain the upper bound of the system delay according to the service capability and load; According to the upper bound of the delay, the online service quality is evaluated. 2. The guarantee performance evaluation method for online quality of service according to claim 1, wherein the analyzing the network performance comprises: analyzing the blocking probability, analyzing the immediate service probability, analyzing the delay time, analyzing the service Concurrency capability, analytical load. 3. A kind of online quality of service guarantee performance evaluation method according to claim 1, is characterized in that, described according to described service capability and load, obtain the delay upper bound of the system specifically: The service capability is β _e2e , the load is a(t), and the upper bound of the delay D _max The upper bound of the delay of the system is obtained by the following formula: Formula 1:

Formula 2: D _max =sup{inf{τ≥0:α(s)≤β _e2e (s+τ)}}; Formula 3:

Among them, R _n1 is the service rate provided by the forward network service, R _n2 is the service rate provided by the backward network service, R _c is the service rate provided by the cloud service, and f is the impact of the cloud service on data transmission when processing the computing process factor. 4. A kind of guarantee performance evaluation method of online service quality according to claim 3, it is characterised in that the online service quality is evaluated according to the time delay upper bound, specifically: Calculating the formulas 1, 2, and 3 to obtain:

Wherein, R _e2e =min{R _n1 ,R _c ,R _n2 /f}, T _e2e =T _n1 +T _c +T _n2 ; When the load is α(t)=M+pt, the upper bound of the system delay is D _max =T _e2e +M/R _e2e =T _c +T _e +M/R _e2e ; Wherein, T _n =T _n1 +T _n2 . 5. The online quality of service assurance performance evaluation method according to claim 4, further comprising: under the condition of ignoring signal processing delay, the service delay parameter T link transmission delay and packet processing The sum of the delays is T=L/R+L/C, where L represents the maximum packet length, C represents the minimum link transmission rate, and R represents the number of user requests. Wherein, T _n =L(1/R _n1 +1/R _n2 +2/C); The upper bound of the delay of the system is D _max =T _c +L(1/R _n1 +1/R _n2 +2/C)+M/R _e2e ; where, R _e2e =min{R _n1 ,R _c ,R _n2 /f}. 6. An online service quality assurance performance evaluation device, characterized in that, comprising: The network performance analysis module is used to analyze the network performance and obtain the service capability of the system and the load of the system; a delay upper bound obtaining module, used for obtaining the system delay upper bound according to the service capability and load; The online service quality evaluation module is used to evaluate the online service quality according to the upper bound of the delay. 7. A kind of online quality of service guarantee performance evaluation device according to claim 6, is characterized in that, described delay upper bound acquisition module is specifically used for: The service capability is β _e2e , the load is a(t), and the upper bound of the delay D _max The upper bound of the delay of the system is obtained by the following formula: Formula 1:

Formula 2: D _max =sup{inf{τ≥0:α(s)≤β _e2e (s+τ)}}; Formula 3:

Among them, R _n1 is the service rate provided by the forward network service, R _n2 is the service rate provided by the backward network service, R _c is the service rate provided by the cloud service, and f is the impact of the cloud service on data transmission when processing the computing process factor. 8. The online service quality assurance performance evaluation device according to claim 7, wherein the online service quality evaluation module is specifically used for: Calculating the formulas 1, 2, and 3 to obtain:

Wherein, R _e2e =min{R _n1 ,R _c ,R _n2 /f}, T _e2e =T _n1 +T _c +T _n2 ; When the load is α(t)=M+pt, the upper bound of the system delay is D _max =T _e2e +M/R _e2e =T _c +T _e +M/R _e2e ; Wherein, T _n =T _n1 +T _n2 . 9 . The online service quality assurance performance evaluation device according to claim 8 , wherein the online service quality evaluation module is further configured to: under the condition of ignoring signal processing delay, the service delay The parameter T is the sum of the link transmission delay and the packet processing delay, that is, T=L/R+L/C, where L represents the maximum packet length, C represents the minimum link transmission rate, and R represents the number of user requests; Wherein, T _n =L(1/R _n1 +1/R _n2 +2/C); The upper bound of the delay of the system is D _max =T _c +L(1/R _n1 +1/R _n2 +2/C)+M/R _e2e ; where, R _e2e =min{R _n1 ,R _c ,R _n2 /f}. 10. An online service quality assurance performance evaluation device, characterized by comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, the processor executing the The computer program implements the online service quality assurance performance evaluation method according to any one of claims 1 to 5.

Images